期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2015
卷号：3
期号：11
DOI：10.15680/IJIRCCE.2015.0311177
出版社：S&S Publications
摘要：Back–propagation is very good and effective method for learning neural networking. This method iswidely used various application. As compare to learning with local data set the collaborative learning is very effectiveto learn new things. The collaborative infrastructure like cloud computing the participating parties carries out learningnot only their own data set, but also on other data set .The cloud computing is more convenient than ever for useracross the internet. The user internet can shared all data without knowing with each other.Beside of this advantage there one crucial issue pertaining to the internet-wide collaborative neural network learning isthe protection of the data privacy for each participant. The participant in the cloud computing is from different trustdomain and they may not Want to release their private data set which may contain proprietary information to anybodyelse. The solution shall be effective and scalable enough to support a random number of participants each processingarbitrarily participant data set.Co-operative data sharing is an important aspect that is emerging heavily in cloud computing. This comes with a largerisk of data leakage from the cloud. Thus a need for encoding data before storing on cloud becomes almost mandatory.But, in a multi owner data access structure it is important for clients to find mining results, to make any system functioneffortlessly. This paper supports architecture for storing data in an encoded manner on the server, yet making it feasibleto apply ANN for mining on encoded data with an encoded query, which makes it impossible for a curious cloud ownerto find and meaning of data, not even from the query.
